Icon

Data Final Project TESKE

Initial Diagnosis

Baseline Models

Advanced Modeling

Data Prep.

Looking for ARIMA selection

ARIMA

ETS

Initial ACF + PACF interpretation:

The slowly decaying ACF, combined with a PACF that cuts off at lag 1, suggests a non-stationary series with strong persistence, consistent with a random walk or a series that requires first differencing.

Final Forecast

Data from: https://www.tsa.gov/travel/passenger-volumes
CSV Reader
To match R code, replaced "number" to passengers
Column Renamer
removed commasfrom passengers
String Manipulation
Converted to Numeric(# of passengers)
String to Number
String to Date&Time
trying to do a R snippet after string to number did not work.Dumb move by me I figured out dates.
Date Rounder
Now R and KNIMEwill get along
GroupBy
Train/Test Split
R Snippet
ETS Model(Train → Forecast Test
R Snippet
ETS Forecast vs Actual
R View (Table)
Auto ARIMA
R Snippet
Final ARIMA Forecast
R Snippet
ETS accuracy
R Snippet
Residual Diagnostics
R Snippet
Train/Test Split
R Snippet
Train/Test Split
R Snippet
ARIMA Family Comparison
R Snippet
Forecast Plot
R View (Table)
ADF Test
R Snippet
PACF
R View (Table)
NAIVE model(Train → Forecast Test)
R Snippet
Drift Model(Train → Forecast Test)
R Snippet
Baseline Accuracy & Residuals
R Snippet
NAIVE forecast plot
R View (Table)
STL Decomp
R Snippet
Drift forecast plot
R View (Table)
Baseline Accuracy & Residuals
R Snippet
R View (Table)
Baseline Accuracy
R Snippet
Time Series Plot
R Snippet
R View (Table)
Residual Diagnostics
R Snippet
ACF
R Snippet
Train/Test Split
R Snippet
R View (Table)
ARIMA model (Train → Forecast Test)
R Snippet
ARIMA Forecast vs Actual
R View (Table)

Nodes

Extensions

Links